METHOD FOR DENOISING UNDERWATER ACOUSTIC SIGNAL ON THE BASIS OF ADAPTIVE WINDOW FILTERING AND WAVELET THRESHOLD OPTIMIZATION

摘要

Provided is a method for denoising an underwater acoustic signal on the basis of adaptive window filtering and wavelet threshold optimization. Said method comprises: first describing, in conjunction with an SαS distribution and normal distribution model, a Gaussian/non-Gaussian pulse noise in an underwater acoustic channel; designing a median filtering method based on an adaptive window, and correcting the size of a filtering window according to the number of noise points in the window, so as to suppress the non-Gaussian pulse noise; and then, on the basis of an improved artificial bee colony method (GDES-ABC), optimizing threshold parameters of a wavelet threshold denoising method, thereby improving the capability of suppressing Gaussian noise. Said method can effectively suppress Gaussian/non-Gaussian pulse noises in a complex underwater acoustic environment, improve the receiving capability of underwater acoustic communication signals such as 2FSK, QPSK and 16QAM, and obtain a higher output signal-to-noise ratio and noise suppression ratio.